Simplified data transition tracking loop—a new NRZ maximum likelihood symbol synchroniser

The True Performance of the Simplified Data Transition Tracking Loop

“…It is the presence of noise in the error signal of the SDTTL during the no-transition intervals (half of the time for ideal detection) compared with a zero error signal of the DTTL in these same intervals that explains the asymptotic (in the limit of a large symbol SNR) 3-dB-worse performance of the former relative to the latter. The authors of [1] completely missed this all-important point in concluding that the two schemes had identical noise behavior.…”

“…About 10 years ago, Chiu and Lee [1] reported on a simplified data-transition tracking loop (SDTTL) which, by using an alternate approximation to the derivative of the rectangular pulse shape, suggested a structure that was able to eliminate the need for the in-phase-arm transition detector and quadraturephase-arm delay element. Based on a limited analysis, i.e., illustration of the noise-free timing waveforms distributed throughout the loop and derivation of the associated noise-free S-curve, these authors claimed that the two structures had exactly equivalent S-curve and noise performances.…”